Effects of Alumina Phase and Loading Amount on Catalytic Methane Combustion Activity of Pd- and Pt-Based Catalysts Kraikul, N., and Jitkarnka, S.*

The Petroleum and Petrochemical College, Chulalongkorn University, Bangkok, Thailand, 10330

© 2004, N. Kraikul and S. Jitkarnka, Prepared for Presentation at 2004 Annual Meething/ November 7-12/ Topics in Fuel Cell Technology, Unpublished, AIChE Shall Not Be Responsible For Statements or Opinions Contained in Papers or Printed in its Publications Abstract

Catalytic combustion of natural gas has been considered as an ongoing challenge to achieve ultra low emissions from gas turbine engine over the past decade. However, this technology still has some problems about the combustion catalyst that is not active at the compressor discharge temperature, and rapidly deactivates at high sufficient temperatures. Therefore, palladium based catalysts were employed in this study to investigate the effect of calcination steps, and the effect of loading amount on gamma-alumina washcoated ceramic monolith. All catalysts were prepared by incipient wetness method, and were screened their catalytic activity and selectivity by a multi-flow reactor equipped with gas chromatography (GC) based on the concept of high throughput screening. The screening results showed that all palladium based catalysts has high potential to combust a mixture of 2%methane in air, and did not be affected by the calcination steps. Five percent of palladium was found to be an optimum loading in order to ensure high combustion activity. Keywords: catalytic combustion, high throughput, alumina phase, loading amount Introduction

Catalytic combustion of natural gas has been considered as an ongoing challenge to achieve ultra low emissions from gas turbine engine over the past decade. There are many advantages promising this technology, particularly for its ability to substantially diminish the amount of exhausting NOX during operations. One way to integrate catalysts into the combustion chamber is placing two catalytic zones (low- and mid-temperature catalytic sections) prior the homogeneous combustion section to avoid the rapid deactivation of catalysts with the high combustion temperature [Dalla Betta and Rostrup-Nielsen, 1999; Forzatti, 2000; Corroni et al., 2002; Thevenin, 2002]. Although the using of catalysts in this design, which is called hybrid design, can prevent the major deactivation of the catalysts, some catalysis solutions still need to be approached to overcome the other obstructing problems and make the design commercialization [Etemad et al., 1999; Ersson, 2003]. As a result, reducing the light-off temperature, maintaining the time-on-stream properties and reducing the catalyst cost associated with the large quantities of noble metal used (~10%wt) of the low temperature catalytic section are particularly important to be achieved.

Recently, using multi-components catalysts has been intensively employed in many kinds of applications due to its promising of more advantages during operations. At the same time, the enhancements of combustion activity upon using Pd/Al2O3 with other metals were also frequency observed, such as Pd-Pt/Al2O3 [Yamamoto and Uchida, 1998; Narui et al., 1999; Janbey, 2003], Pd-La/Al2O3 [Kennelly et al., 1993; Ozawa et al., 2003(a)], Pd-La-Nd/Al2O3 [Ozawa et al., 2003(b)] and Pd-La-Ce/Al2O3 [Thevenin, 2003]. These results implied that the use of multi-element catalysts appears to enhance some properties of the combustion catalyst, and is an interesting system to study further.

Even though the bimetallic combination of Pd with Pt, and with La had already been studied in the past, as previously mentioned, the combination of using these three elements together is still rarely reported [Ozawa et al., 2004]. Therefore, the study of Pd-Pt-La/Al2O3 catalyst is still very challenging, especially in more details among their various compositions. However, before the study on in Pd-Pt-La tri-element system is conducted, the preparation parameters that could affect to the combustion activity were prior investigated to evaluate the best conditions for preparing catalysts in the further tri-element study.

In this contribution, two parameters affecting on combustion activity of palladium based catalysts, including calcination steps relating to alumina phase and loading amount on �-Al2O3 washcoated ceramic monolith, were investigated. A homemade eight-tubular flow reactor further modified from the literature [Pérez-Ramírez et al., 2000] based on high throughput screening concept was employed to complete the activity tests in rapid fashion. XRD and SEM technique were also performed on those studies for observing some characteristics of the catalysts. Experimental

1. Catalysts Preparation

Washcoated ceramic monoliths prepared from dipping 2�2�15 mm3 monolith into 40% (w/v) acidic �-Al2O3 (Alfa® AESAR®) slurry with coating thickness around 30-35 �m (confirmed by SEM) were used as a solid support. Pd-based catalysts were prepared by using Pd(NO3)2�2H2O (Fluka, Switzerland), H2PtCl6�6H2O (Carlo Erba, Italy), and La(NO3)3�6H2O (Fluka, Switzerland) as the active element precursors. Two catalyst libraries were prepared for studying the effects of calcination steps relating to alumina phase and loading amount as hereinbefore mentioned.

For the study on the effect of calcination steps relating to alumina phase, two different calcination procedures related to two alumina phases, �-Al2O3 and �-Al2O3, were employed for preparing catalysts as listed in Table 1. There are representatives of catalysts in our target groups expected for further studies. The purpose of the first calcination was to strengthen the washcoated Al2O3 onto the monolithic substrate while the second calcination was conducted to decompose the active elements precursors. For Procedure A, washcoated monoliths were dried overnight and calcined at 500�C for 3 hrs (yielding �-Al2O3) before being impregnated with a mixed Pd-based or Pt-based solution of active elements to achieve 5% of total loading. After that, they were dried overnight and re-calcined at 900�C with the same holding time before being investigated for their catalytic activity. The Procedure B was quite the same, except the temperature in the first calcination step being changed to 900�C (creating �-Al2O3). The phase transformation of �-Al2O3 due to calcinations were monitored by XRD. The catalytic activity results gave a decision to select the better procedure for the next study on effect of loading amount.

Another eight catalysts with various loading amounts of Pd in the range of 1-12%wt were prepared by incipient wetness method following the selected calcination procedure from the previous experiments for studying the effect of loading amount. The combustion activity results also gave a decision to select the best loading amount for the further investigation on tri-element system.

Table 1 Catalysts formulations for studying the effect of Al2O3 phase

Relative ratio of elemental loading† (%) Catalyst Pd Pt La

Remark

BM - - - Bare Monolith WM - - - Washcoated Monolith Pd80-Pt20 (A) 80 20 - Procedure A (Pd-based) Pd80-Pt20 (B) 80 20 - Procedure B (Pd-based) Pd60-Pt20-La20 (A) 60 20 20 Procedure A (Pd-based) Pd60-Pt20-La20 (B) 60 20 20 Procedure B (Pd-based) Pt80-La20 (A) - 80 20 Procedure A (Pt-based) Pt80-La20 (B) - 80 20 Procedure B (Pt-based)

Note: †with 5% total elemental loading

2. Catalytic Activity Measurements

All combustion experiments were performed using the homemade eight-tubular flow reactor (Figure 1) with monolithic catalysts placed inside each ¼” stainless U-tube reactor submerged into a furnace. The temperature of furnace was controlled by Maxthermo temperature controller (MC2438) equipped with K-type thermocouple. Purging nitrogen was also applied to ensure a good distribution of temperature inside the furnace. Prior to the combustion study, H2 was purged throughout all reactors at around 600�C for one hour, aiming to achieve a metallic state of some active elements before being screened for their combustion activity. Afterwards, N2 was switched to flush the remaining H2 and avoid the local temperature rise during the change of temperature set point. A lean mixture containing 2% methane (99.99% purity) in air with 1 m s-1 linear velocity was used as the feed to investigate combustion activity under 1 atmospheric pressure, and selected temperatures. This testing condition was simulated from the operating condition of a small-size gas turbine engine. The flue gases were passed to water trapper before being sent to a multi-channel valve in order to control the stream traffic to gas chromatograph (GC). Only one stream at a time was sent to analyze in HP 6890N GC using HP-PLOT Q and HP-MOLSIV columns equipped with thermal conductivity detector. After a sampling gas was passed into GC column, the multi-channel valve selected another stream and sent to GC, serving for the next analyzing cycle. The cycle was repeated until all of the eight streams were analyzed along time-on-stream.

3. Catalysts Characterization

3.1 Scanning Electron Microscope (SEM): The unique information about washcoated thickness was confirmed by scanning electron microscope using a JEOL 5220 instrument. Samples were stuck on the stubs and coated with gold by ion sputtering device (JFC-1100E) for 4 min prior being examined for their surface with a magnification ranges from 50 to 700.

3.2 X-ray Diffraction (XRD): The change of alumina phase due to different calcination procedures were monitored by using X-ray powder diffraction spectroscopy (XRD), performed on the Rigaku X-ray diffractometer system with a Rint 200 wide-angle goniometer, Cu-K� radiation, 30 mA and 40 kV. Scan speed of 5� (2�)/min with a scan step of 0.02� (2�) was applied during a continuous run in the 32� to 48� (2�) range. Phase identification was carried out using the reference database (JCPDS-files) supplied with the equipment.

Figure 1 Schematic flow diagram of the homemade eight tubular flow reactors equipped with gas chromatograph (GC)

Results and Discussions

1. SEM of Washcoated Monolith The thickness of �-Al2O3 washcoating layer was estimated using SEM after dipping monolith into acidic �-Al2O3 slurry and calcined at 500�C. Figures 2(a) to 2(c) show the SEM images of washcoated monolith, which indicate the thickness of around 30-35 �m coated layer compared with that image of bare monolith, as shown in Figure 2(d).

1.2 XRD Patterns of Al2O3 from Different Calcination Procedures The XRD patterns of two calcined Al2O3 from the different calcination procedures are shown in Figure 3. The results indicated that two different alumina phases were obtained from the differences in calcination procedure: �-Al2O3 from Procedure A and mixed of �- and �-Al2O3 phases from Procedure B. These two different phases of alumina were loaded with the same active elements compositions for investigating their effect on the combustion activity in the next catalytic activity testing.

Figure 2 SEM images of �-Al2O3 washcoated monolith, with the magnifications of (a) 50, (b) 75, (c) 750 times, and (d) bare monolith Figure 3 XRD patterns of original �-Al2O3, calcined �-Al2O3 at 500 and 900�C for 3 hours.

3. Catalytic Activity Measurements Figure 4 shows the combustion activity results from screenings of the listed catalysts in Table 1 at 650°C. It was found that, the combustion activity did not significantly affect when the loading on different alumina phases were applied to prepare Pd-based catalysts [Pd80-Pt20 (A), Pd80-Pt20 (B), Pd60-Pt20-La20 (A), and Pd60-Pt20-La20 (B)]. On the other hand, loading Pt-based elements on �-phase alumina, according to preparation Procedure A [Pt80-La20 (A)], gave a higher activity when compared to the other one from Procedure B [Pt80-La20 (A)]. The results indicated that the activity of Pt-based catalyst (in absence of Pd) is susceptible to the change of Al2O3 phase, related to the reduction of surface area. Thus, Procedure A, which washcoated monoliths were pre-calcined at 500�C before re-calcination at 900�C after metal loading, was chosen to be the

�-Al2O3 coated layer

ceramic monolith

(a) (b) (c) (d)

Coated Layer

Ceramic Monolith

5 15 25 35 45 55 65 75 85

2����

Inte

nsity

Original �-Al2O3

Calcined at 500�C

Calcined at 900�C

�

� � �

�

� � �

�

� �

� �

� �

� BM � WM � Pd80-Pt20 (A) � Pd80-Pt20 (B)

� Pd60-Pt20-La20 (A) � Pd60-Pt20-La20 (B) � Pt80-La20 (A) � Pt80-La20 (B)

better condition for the further preparation of methane combustion catalysts in further studies.

For the study on the effect of loading amount, eight prepared catalysts with different Pd loading amounts in the range of 1-12%wt were investigated for their combustion activity under three different temperatures: 350, 400, and 450ºC, as shown in Figure 5. Figure 5a representing the methane conversion at 350�C shows the needs of palladium to be loaded in higher amounts at this temperature since the activity of all catalysts is still low, around 10-15% conversion. However, the results at 450�C, as illustrated in Figure 5c, evidenced that it was not necessary to load metals in very high amounts because most of catalysts with about 4% total loading can give almost complete combustion of methane at this temperature. In Figure 5b, the methane conversion at 400�C is exhibited, and also confirms the conclusion that it was not necessary to load metals in very high amounts since almost the same activity, around 25-35% conversion, were observed for all catalysts containing about 4% total loading or higher in a long period of time. Therefore, 5% is suggested to be an optimum total loading on the accounts of high activity assured and metals cost benefited.

Figure 4 Methane conversion over the different catalysts prepared from the different calcination procedure.

Conclusions For the study of two parameters affecting to combustion activity of Pd-based catalysts, the suitable preparation conditions have been suggested from the combustion activity screenings in this work for future studies. Even though the combustion activity of all Pd-based catalysts did not affect by the phase of alumina, it was recommended to prepare the catalysts following Procedure A if in multi-component studies, the catalysts consist of Pt as a component, because the activity of Pt-based catalysts (in absence of Pd) is susceptible to the phase change of alumina. Furthermore, 5%wt loading of Pd was also suggested to be an optimum loading since the high combustion activity can be ensured, and the cost associated with the quantity of noble metal used still not too much until it becomes trouble.

Figure 5 Methane conversion over the catalysts with various Pd loading amounts at (a) 350, (b) 400, and (c) 450°C

Acknowledgements

This work was partially funded by Postgraduate Education and Research Programs in Petroleum and Petrochemical Technology (PPT Consortium under ADB fund). The authors would like to thank Prof. Daniel E. Resasco for his fruitful discussions and for providing us the bare monolith.

Literatures Cited Carroni, R., Schmidt, V., and Griffin T., Catalytic combustion for power generation, Catal.

Today, 75, 287-295, 2002. Dalla Betta, R.A. and Rostrup-Nielsen, T., Application of catalytic combustion to a 1.5 MW

industrial gas turbine, Catal. Today, 47, 369-375, 1999. Ersson, A., Material for High-Temperature Catalytic Combustion. Ph.D. Thesis in chemical

engineering, Department of Chemical Engineering and Technology, Kingliga Tekniska Högskolan (Royal Institute of Technology), 2003.

Etemad, S., Karim, H., Smith, L. L., and Pfefferle., Advance technology catalytic combustior for high temperature ground power gas turbine applications, Catal. Today, 47, 305-313, 1999.

Forzatti, P., Environmental catalysis for stationary applications, Catal. Today, 62, 50-65, 2000.

Janbey, A., Clark, W., Noordally, E., Grimes, S., Tahir, S., Noble metal catalysts for methane removal. Chemosphere, 52, 1041–1046, 2003.

Kennelly, J., Mead, B., and Freeauto, R. J., Catalytic combustion process using supported palladium oxide catalysts, US Pat. 5 216 875, 1993.

Narui, K., Yata, H., Furuta, K., Nishida, A., Kohtoku, Y. and Matsuzaki, T., Effects of adding of Pt to PdO/Al2O3 catalyst on catalytic activity for methane combustion and TEM observations of supported particles, Appl. Catal. A, 179, 165-173, 1999.

Ozawa, Y., Tochihara, Y., Watanabe, A., Nagai. M., and Omi, S., Stabilizing effect of Nd2O3, La2O3 and ZrO2 on Pt·PdO/Al2O3 during catalytic combustion of methane, Appl. Catal. A, 258, 261-267, 2004.

Ozawa, Y., Tochihara, Y., Nagai, M., and Omi, S., PdO/Al2O3 in catalytic combustion of methane: stabilization and deactivation, Chem. Eng. Sci., 58, 671-677, 2003(a).

Ozawa, Y., Tochihara, Y., Nagai, M., and Omi, S, Effect of addition of Nd2O3 and La2O3 to PdO/Al2O3 in catalytic combustion of methane, Catal. Commun., 4, 87-90, 2003(b).

Pérez-Ramírez, J., Berger, R., Mul, G., Kapteijn, F., and Moulijn, J. A, The six-flow reactor technology: a review on fast catalyst screening and kinetic studies, Catal. Today, 60, 93-109, 2000.

Thevenin, P.O., Alcalde, A., Pettersson, L.J., Järås, S.G., and Fierro, J.L.G., Catalytic combustion of methane over cerium-doped palladium catalysts, J. Catal., 215, 78-86, 2003.

Thevenin, P., Catalytic Combustion of Methane. Ph.D. Thesis in chemical engineering, Department of Chemical Engineering and Technology, Kingliga Tekniska Högskolan (Royal Institute of Technology), 2002.

Yamamoto, H. and Uchida, H., Oxidation of methane over Pt and Pd supported on alumina in lean-burn natural-gas engine exhaust, Catal. Today, 45, 147-151, 1998.